U.S. patent application number 10/214852 was filed with the patent office on 2003-10-02 for estimating traffic distribution in a mobile communication network.
Invention is credited to Friydin, Boris, Shafran, Gil.
Application Number | 20030186693 10/214852 |
Document ID | / |
Family ID | 28456818 |
Filed Date | 2003-10-02 |
United States Patent
Application |
20030186693 |
Kind Code |
A1 |
Shafran, Gil ; et
al. |
October 2, 2003 |
Estimating traffic distribution in a mobile communication
network
Abstract
A method for estimating traffic distribution in a mobile
communication network includes collecting statistical information
with regard to a quantity of communication traffic and with regard
to a quality indicator associated with the traffic in a region
served by the mobile communication network. The region is divided
into areas belonging to respective traffic types. A respective
traffic density is estimated for each of the traffic types based on
the statistical information collected with regard to the quantity
of the traffic and the quality indicator.
Inventors: |
Shafran, Gil; (Jerusalem,
IL) ; Friydin, Boris; (Rehovot, IL) |
Correspondence
Address: |
BLAKELY SOKOLOFF TAYLOR & ZAFMAN
12400 WILSHIRE BOULEVARD, SEVENTH FLOOR
LOS ANGELES
CA
90025
US
|
Family ID: |
28456818 |
Appl. No.: |
10/214852 |
Filed: |
August 7, 2002 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60369368 |
Apr 1, 2002 |
|
|
|
Current U.S.
Class: |
455/423 ;
455/446 |
Current CPC
Class: |
H04W 24/00 20130101;
H04W 16/18 20130101 |
Class at
Publication: |
455/423 ;
455/446 |
International
Class: |
H04Q 007/20 |
Claims
1. A method for estimating traffic distribution in a mobile
communication network, comprising: collecting statistical
information with regard to a quantity of communication traffic and
with regard to a quality indicator associated with the traffic in a
region served by the mobile communication network; dividing the
region into areas belonging to respective traffic types; and
estimating a respective traffic density for each of the traffic
types based on the statistical information collected with regard to
the quantity of the traffic and the quality indicator.
2. A method according to claim 1, wherein the network comprises a
plurality of fixed transceivers at respective locations in the
region, and wherein collecting the statistical information
comprises collecting the information from the fixed transceivers
with respect to the communication traffic exchanged over the air
between the fixed transceivers and mobile units served by the
network.
3. A method according to claim 2, wherein the network comprises a
cellular network, and wherein collecting the information from the
fixed transceivers comprises collecting the information with
respect to cells in the network that are served by the fixed
transceivers.
4. A method according to claim 3, wherein collecting the
information with regard to the quality indicator comprises
collecting statistics regarding handoffs between the cells.
5. A method according to claim 3, wherein dividing the region
comprises dividing the region into bins, associating the bins with
respective clutter types, and defining each of the traffic types by
grouping together all the bins that belong a respective one of the
clutter types and are all served by a respective one of the
cells.
6. A method according to claim 2, and comprising measuring time
delays in transmission of the communication traffic between the
fixed transceivers and the mobile units, and wherein estimating the
respective traffic density comprises using the time delays in
determining the traffic density.
7. A method according to claim 2, wherein collecting the
information comprises measuring an effect of interference by a
first one of the fixed transceivers on the traffic exchanged
between the mobile units and a second one of the fixed
transceivers.
8. A method according to claim 7, wherein measuring the effect
comprises collecting statistics regarding carrier/interference
values in the traffic exchanged between the mobile units and the
second one of the fixed transceivers.
9. A method according to claim 7, wherein measuring the effect
comprises determining an element of an impact matrix relating the
first and second ones of the fixed transceivers.
10. A method according to claim 7, wherein measuring the effect
comprises collecting statistics regarding dropped call rates.
11. A method according to claim 2, and comprising optimizing a
configuration of the fixed transceivers responsive to the estimated
traffic density.
12. A method according to claim 11, wherein optimizing the
configuration comprises distributing operating frequencies among
the fixed transceivers responsive to the estimated traffic
density.
13. A method according to claim 1, wherein collecting the
statistical information with regard to the quality indicator
comprises collecting statistics with regard to a signal/noise ratio
associated with the traffic.
14. A method according to claim 1, wherein collecting the
statistical information with regard to the quality indicator
comprises collecting statistics with regard to a power level of
received signals used in carrying the traffic.
15. A method according to claim 1, wherein dividing the region
comprises dividing the region into bins, associating the bins with
respective clutter types, and defining each of the traffic types by
grouping together all the bins in mutual proximity that belong a
respective one of the clutter types.
16. A method according to claim 1, wherein dividing the region
comprises defining the areas in accordance with a grid imposed on
the region.
17. A method according to claim 1, wherein the communication
traffic comprises voice traffic.
18. A method according to claim 1, wherein the communication
traffic comprises packet data traffic.
19. Apparatus for estimating traffic distribution in a mobile
communication network, comprising a computer, which is coupled to
collect statistical information with regard to a quantity of
communication traffic and with regard to a quality indicator
associated with the traffic in a region served by the mobile
communication network, wherein the region is divided into areas
belonging to respective traffic types, the computer is adapted to
estimate a respective traffic density for each of the traffic types
based on the statistical information collected with regard to the
quantity of the traffic and the quality indicator.
20. Apparatus according to claim 19, wherein the network comprises
a plurality of fixed transceivers at respective locations in the
region, and wherein the statistical information is provided by the
fixed transceivers with respect to the communication traffic
exchanged over the air between the fixed transceivers and mobile
units served by the network.
21. Apparatus according to claim 20, wherein the network comprises
a cellular network, and wherein the statistical information is
provided with respect to cells in the network that are served by
the fixed transceivers.
22. Apparatus according to claim 21, wherein the statistical
information comprises statistics regarding handoffs between the
cells.
23. Apparatus according to claim 21, wherein the region is divided
into bins, which are associated with respective clutter types, and
wherein each of the traffic types is defined by grouping together
all the bins that belong a respective one of the clutter types and
are all served by a respective one of the cells.
24. Apparatus according to claim 20, wherein the computer is
adapted to receive measurements of time delays in transmission of
the communication traffic between the fixed transceivers and the
mobile units, and to estimate the respective traffic density using
the time delays.
25. Apparatus according to claim 20, wherein the statistical
information comprises statistics regarding an effect of
interference by a first one of the fixed transceivers on the
traffic exchanged between the mobile units and a second one of the
fixed transceivers.
26. Apparatus according to claim 25, wherein the statistics
comprise data regarding carrier/interference values in the traffic
exchanged between the mobile units and the second one of the fixed
transceivers.
27. Apparatus according to claim 25, wherein the computer is
adapted to determine, responsive to the statistics, an element of
an impact matrix relating the first and second ones of the fixed
transceivers.
28. Apparatus according to claim 25, wherein the statistics
comprise data regarding dropped call rates.
29. Apparatus according to claim 20, wherein the computer is
adapted to determine an optimized configuration of the fixed
transceivers responsive to the estimated traffic density.
30. Apparatus according to claim 29, wherein the optimized
configuration comprises an optimized distribution of operating
frequencies among the fixed transceivers based on the estimated
traffic density.
31. Apparatus according to claim 19, wherein the statistical
information with regard to the quality indicator comprises
statistics with regard to a signal/noise ratio associated with the
traffic.
32. Apparatus according to claim 19, wherein the statistical
information with regard to the quality indicator comprises
statistics with regard to a power level of received signals used in
carrying the traffic.
33. Apparatus according to claim 19, wherein the region is divided
into bins, the bins are associated with respective clutter types,
and each of the traffic types is defined by grouping together all
the bins in mutual proximity that belong a respective one of the
clutter types.
34. Apparatus according to claim 19, wherein the region is divided
into the areas in accordance with a grid imposed on the region.
35. Apparatus according to claim 19, wherein the communication
traffic comprises voice traffic.
36. Apparatus according to claim 19, wherein the communication
traffic comprises packet data traffic.
37. A computer software product for estimating traffic distribution
in a mobile communication network, the product comprising a
computer-readable medium in which program instructions are stored,
which instructions, when read by a computer, cause the computer to
receive statistical information collected with regard to a quantity
of communication traffic and with regard to a quality indicator
associated with the traffic in a region served by the mobile
communication network, wherein the region is divided into areas
belonging to respective traffic types, and wherein the instructions
cause the computer to estimate a respective traffic density for
each of the traffic types based on the statistical information
collected with regard to the quantity of the traffic and the
quality indicator.
38. A product according to claim 37, wherein the network comprises
a plurality of fixed transceivers at respective locations in the
region, and wherein the statistical information is provided by the
fixed transceivers with respect to the communication traffic
exchanged over the air between the fixed transceivers and mobile
units served by the network.
39. A product according to claim 38, wherein the network comprises
a cellular network, and wherein the statistical information is
provided with respect to cells in the network that are served by
the fixed transceivers.
40. A product according to claim 39, wherein the statistical
information comprises statistics regarding handoffs between the
cells.
41. A product according to claim 39, wherein the region is divided
into bins, which are associated with respective clutter types, and
wherein each of the traffic types is defined by grouping together
all the bins that belong a respective one of the clutter types and
are all served by a respective one of the cells.
42. A product according to claim 38, wherein the instructions cause
the computer to receive measurements of time delays in transmission
of the communication traffic between the fixed transceivers and the
mobile units, and to estimate the respective traffic density using
the time delays.
43. A product according to claim 38, wherein the statistical
information comprises statistics regarding an effect of
interference by a first one of the fixed transceivers on the
traffic exchanged between the mobile units and a second one of the
fixed transceivers.
44. A product according to claim 43, wherein the statistics
comprise data regarding carrier/interference values in the traffic
exchanged between the mobile units and the second one of the fixed
transceivers.
45. A product according to claim 43, wherein the instructions cause
the computer to determine, responsive to the statistics, an element
of an impact matrix relating the first and second ones of the fixed
transceivers.
46. Apparatus according to claim 43, wherein the statistics
comprise data regarding dropped call rates.
47. A product according to claim 38, wherein the instructions cause
the computer to determine an optimized configuration of the fixed
transceivers responsive to the estimated traffic density.
48. A product according to claim 47, wherein the optimized
configuration comprises an optimized distribution of operating
frequencies among the fixed transceivers based on the estimated
traffic density.
49. A product according to claim 38, wherein the statistical
information with regard to the quality indicator comprises
statistics with regard to a signal/noise ratio associated with the
traffic.
50. A product according to claim 38, wherein the statistical
information with regard to the quality indicator comprises
statistics with regard to a power level of received signals used in
carrying the traffic.
51. A product according to claim 38; wherein the region is divided
into bins, the bins are associated with respective clutter types,
and each of the traffic types is defined by grouping together all
the bins in mutual proximity that belong a respective one of the
clutter types.
52. A product according to claim 38, wherein the region is divided
into the areas in accordance with a grid imposed on the region.
53. A product according to claim 38, wherein the communication
traffic comprises voice traffic.
54. A product according to claim 38, wherein the communication
traffic comprises packet data traffic.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application claims the benefit of U.S. provisional
patent application No. 60/369,368, filed Apr. 1, 2002, which is
incorporated herein by reference.
COMPUTER PROGRAM LISTING APPENDIX
[0002] A computer program listing appendix is submitted herewith on
one compact disc and one duplicate compact disc. The total number
of compact discs including duplicates is two. The file on the
compact disc is a Microsoft Excel.RTM. worksheet named
traffDistrib.xls, created Jun. 25, 2002, of length 565,248
bytes.
FIELD OF THE INVENTION
[0003] The present invention relates generally to optimization of
resource use in mobile communication networks, and specifically to
estimation of traffic distribution in such networks.
BACKGROUND OF THE INVENTION
[0004] Service quality in cellular voice networks is typically
measured by a number of key performance indicators:
[0005] System coverage--the geographic extent over which the
network will reliably provide service. This indicator relates not
only to the region over which the network extends, but also the
existence of local coverage "holes."
[0006] Call blockage--the percentage of attempts to make or receive
calls that are blocked due to lack of available voice channels.
Inadequate system capacity leads to high blockage rates.
[0007] Voice quality--the level of noise and/or distortion in voice
conversations, typically measured in terms of bit error rate (BER),
Frame Erasure Rate (FER) and/or Received Level Quality
(RxQual).
[0008] Dropped call rate--percentage of calls in progress that
terminate before either party intentionally ends the call.
[0009] Similar concerns as to coverage, capacity and error rate
exist in wireless packet data networks, although in this case the
key performance indicators relate to whether an online connection
is available to the user and the effective throughput (data rate)
of the connection.
[0010] The key performance indicators are themselves dependent on
characteristics of the underlying radio network that is used to
carry the voice or data signals. Each cell in the network has one
or more antennas that are meant to serve mobile units (cellular
telephones and/or data terminals) within its service area. The
strength of the signals reaching the mobile units from the
antennas, and vice versa, are determined by the path loss of
electromagnetic waves propagating between the antennas and the
mobile unit locations. If the received signal level at a given
location is too low, poor quality or coverage holes will result. In
planning cellular networks, path loss maps are typically used to
locate the antennas and determine the power levels needed to avoid
such holes.
[0011] Each cell in a narrowband cellular network is assigned a
fixed set of frequencies. (Narrowband networks include Time
Division Multiple Access [TDMA] networks, such as Global System for
Mobile [GSM] communication networks. Code Division Multiple Access
[CDMA] networks assign a broad frequency band to each cell.) When a
mobile unit initiates or receives a call, it is assigned to one of
the frequencies of the serving cell. If there is no frequency
available--due typically to traffic in the area of the mobile unit
that is in excess of the capacity of the cell--the call will be
blocked. When a mobile unit, such as a cellular telephone in a car,
moves within the network service region, it may be handed over from
one cell to another. If the new cell does not have a frequency
available, the call will be dropped.
[0012] Thus, in planning the location and configuration of antennas
and the allocation of frequencies in a cellular network, it is
important to take into account the distribution of communication
traffic in every area of the network service region. Each cell
should have sufficient frequency allocation to accommodate the
expected number of mobile units in its service area, so that
blocked and dropped calls are minimized. On the other hand, because
frequency spectrum is a scarce resource in cellular networks,
excess, wasted capacity should be avoided, as well. Therefore,
cellular network operators need accurate traffic distribution
information in order to optimize the key performance indicators of
their networks.
[0013] A number of methods are known in the art for estimating
network traffic distribution. One method is to trace the location
and performance of individual mobile units in the network.
Typically, a small number of special mobile units with geographical
locating capabilities are used for this purpose. Alternatively,
measurements may be made using larger numbers of ordinary mobile
units, by estimating the position of each mobile unit based on
signal strength measurements. In either case, the measurements are
cumbersome and have low statistical reliability.
[0014] An alternative method for estimating traffic distribution is
to use traffic statistics provided by the network itself. The
statistics indicate the amount of traffic served by each cell in
the network during a given measurement period. The statistical
information can be used to estimate the traffic density for each of
a number of different "clutter types" in the service region, such
as urban areas, roads and open space. The problem with this method,
however, is that the granularity of the collected information is
coarse, consisting only of the total traffic per cell. Therefore,
the traffic density calculated in this manner gives only a very
rough estimate of the actual traffic in any particular location in
the network coverage region.
SUMMARY OF THE INVENTION
[0015] It is an object of some aspects of the present invention to
provide improved methods and systems for estimating traffic
distribution in a mobile communication network.
[0016] Cellular networks regularly gather statistical data from
each cell not only on the amount of traffic served, but also with
regard to various indicators of the quality of calls carried by the
cell. These quality indicators include, for example, the average
carrier/interference (C/I) ratio, specific levels of interference
from other cells, and frequency of handovers between cells. In
preferred embodiments of the present invention, these quality
indicators are used together with the measured quantity of traffic
carried by each cell to map the actual traffic density in the
network. The use of the quality statistics in the calculation
allows the network service region to be divided into clutter
classifications with much finer granularity than can be achieved by
methods known in the art. The resulting traffic density map is thus
more accurate and true to reality, allowing better optimization of
the antenna configurations and frequency distribution among the
cells.
[0017] There is therefore provided, in accordance with a preferred
embodiment of the present invention, a method for estimating
traffic distribution in a mobile communication network,
including:
[0018] collecting statistical information with regard to a quantity
of communication traffic and with regard to a quality indicator
associated with the traffic in a region served by the mobile
communication network;
[0019] dividing the region into areas belonging to respective
traffic types; and
[0020] estimating a respective traffic density for each of the
traffic types based on the statistical information collected with
regard to the quantity of the traffic and the quality
indicator.
[0021] Typically, the network includes a plurality of fixed
transceivers at respective locations in the region, and collecting
the statistical information includes collecting the information
from the fixed transceivers with respect to the communication
traffic exchanged over the air between the fixed transceivers and
mobile units served by the network. In a preferred embodiment, the
network includes a cellular network, and collecting the information
from the fixed transceivers includes collecting the information
with respect to cells in the network that are served by the fixed
transceivers. Preferably, collecting the information with regard to
the quality indicator includes collecting statistics regarding
handoffs between the cells. Alternatively or additionally, dividing
the region includes dividing the region into bins, associating the
bins with respective clutter types, and defining each of the
traffic types by grouping together all the bins that belong a
respective one of the clutter types and are all served by a
respective one of the cells.
[0022] In a preferred embodiment, measuring time delays in
transmission of the communication traffic between the fixed
transceivers and the mobile units, and estimating the respective
traffic density includes using the time delays in determining the
traffic density.
[0023] In a further preferred embodiment, collecting the
information includes measuring an effect of interference by a first
one of the fixed transceivers on the traffic exchanged between the
mobile units and a second one of the fixed transceivers.
Preferably, measuring the effect includes collecting statistics
regarding carrier/interference values in the traffic exchanged
between the mobile units and the second one of the fixed
transceivers. Alternatively or additionally, measuring the effect
includes determining an element of an impact matrix relating the
first and second ones of the fixed transceivers. Further
alternatively or additionally, measuring the effect includes
collecting statistics regarding dropped call rates.
[0024] Typically, the method includes optimizing a configuration of
the fixed transceivers responsive to the estimated traffic density.
In a preferred embodiment, optimizing the configuration includes
distributing operating frequencies among the fixed transceivers
responsive to the estimated traffic density.
[0025] Preferably, collecting the statistical information with
regard to the quality indicator includes collecting statistics with
regard to a signal/noise ratio associated with the traffic.
Additionally or alternatively, collecting the statistical
information with regard to the quality indicator includes
collecting statistics with regard to a power level of received
signals used in carrying the traffic.
[0026] Preferably, dividing the region includes dividing the region
into bins, associating the bins with respective clutter types, and
defining each of the traffic types by grouping together all the
bins in mutual proximity that belong a respective one of the
clutter types. Alternatively or additionally, dividing the region
includes defining the areas in accordance with a grid imposed on
the region.
[0027] Typically, the communication traffic includes at least one
of voice traffic and packet data traffic.
[0028] There is also provided, in accordance with a preferred
embodiment of the present invention, apparatus for estimating
traffic distribution in a mobile communication network, including a
computer, which is coupled to collect statistical information with
regard to a quantity of communication traffic and with regard to a
quality indicator associated with the traffic in a region served by
the mobile communication network, wherein the region is divided
into areas belonging to respective traffic types, the computer is
adapted to estimate a respective traffic density for each of the
traffic types based on the statistical information collected with
regard to the quantity of the traffic and the quality
indicator.
[0029] There is additionally provided, in accordance with a
preferred embodiment of the present invention, a computer software
product for estimating traffic distribution in a mobile
communication network, the product including a computer-readable
medium in which program instructions are stored, which
instructions, when read by a computer, cause the computer to
receive statistical information collected with regard to a quantity
of communication traffic and with regard to a quality indicator
associated with the traffic in a region served by the mobile
communication network, wherein the region is divided into areas
belonging to respective traffic types, and wherein the instructions
cause the computer to estimate a respective traffic density for
each of the traffic types based on the statistical information
collected with regard to the quantity of the traffic and the
quality indicator.
[0030] The present invention will be more fully understood from the
following detailed description of the preferred embodiments
thereof, taken together with the drawings in which:
BRIEF DESCRIPTION OF THE DRAWINGS
[0031] FIG. 1 is a schematic, pictorial view of a region served by
a cellular communication network, in accordance with a preferred
embodiment of the present invention; and
[0032] FIG. 2 is a flow chart that schematically illustrates a
method for estimating traffic distribution in a cellular
communication network, in accordance with a preferred embodiment of
the present invention.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
[0033] FIG. 1 is a schematic, pictorial view of a region 20 served
by a cellular communication network, which is optimized in
accordance with a preferred embodiment of the present invention.
For the purposes of the cellular network, region 20 is divided into
partly-overlapping cells, as is known in the art, each served by
one or more fixed transceivers, represented by antennas 22. By way
of example, an antenna 22A serves a cell, which will be referred to
as cell A, in which a mobile unit 23 is being used to carry on a
telephone call. Another antenna 22B serves a neighboring or nearby
cell, which will be referred to as cell B. In the description that
follows, cells A and B will be used to exemplify the possible
influences of one cell (cell B) on the communication quality
experienced by mobile units in another cell (cell A). In the course
of a telephone call, particularly while traveling, mobile unit 23
may be handed off from cell A to cell B, meaning that antenna 22B
serves the mobile unit in place of antenna 22A.
[0034] Region 20 is characterized by a number of different clutter
types, for example, a dense urban area 24, an urban residential
area 26, an industrial area 28, a rural area 30, open space 32 and
a highway 34. Each of these areas, clearly, will have its own
characteristic traffic density. Furthermore, sub-areas within these
predefined clutter types may have their own density
characteristics, depending on the particular nature and uses of the
structures and other features in these sub-areas. Thus, in
principle, each clutter type encountered in region 20 may be broken
into sub-types corresponding to these sub-areas. Preferred
embodiments of the present invention, as described below, provide
methods for defining these sub-types and determining their traffic
density characteristics. The traffic density served by any given
antenna 22 will be a function of the sub-types and sizes of the
sub-areas that fall within the cell served by the particular
antenna.
[0035] Communication traffic in the cellular network serving region
20 is controlled and routed among antennas 22 by a mobile switching
center (MSC) 36, as is known in the art. Typically, the MSC also
collects traffic density and quality statistics from every cell in
region 20. Alternatively, these statistics may be collected by
another management element in the cellular network. Different types
of quality statistics that may be used for the purposes of the
present invention are described below. The traffic density and
quality statistics are passed to a computer 37 for analysis, along
with other information concerning the network configuration. This
other information may include, for example, the configurations of
antennas 22, such as their frequency allocations, locations,
height, transmission power, azimuth and tilt; geographical features
of region 20; and path loss maps, showing the attenuation of
electromagnetic waves propagating between each of the antennas and
different mobile unit locations in region 20.
[0036] Computer 37 processes the per-cell traffic density and
quality statistics for all the cells in region 20 in order to
arrive at a traffic density estimate for each of the clutter
sub-types in the region. To this end, region 20 is divided into
bins 38, each comprising a small geographical area, preferably much
smaller than the size of a cell. Bin sizes may typically be set
between 20.times.20 m and 300.times.300 m, although larger or
smaller bins may also be used, depending on application
requirements. The bins are grouped together into sets corresponding
to different clutter sub-types, and the characteristic sub-type
traffic densities are then estimated, in a manner described below.
The computer performs these functions under the control of software
supplied for this purpose. The software may be conveyed to the
computer in electronic form, over a network, for example, or it may
be furnished on tangible media, such as CD-ROM.
[0037] FIG. 2 is a flow chart that schematically illustrates a
method for estimating the traffic density by sub-type in region 20,
in accordance with a preferred embodiment of the present invention.
For each cell in the region, computer 37 receives a measure of the
traffic density in that cell, at a traffic measurement step 40. The
traffic density is typically expressed in units of Erlangs,
corresponding to one hour of call time per temporal hour. For any
given cell, say cell A, the total traffic density T(A) is given by:
1 T ( A ) = x X T ( x ) p ( S ( A , x ) ) ( 1 )
[0038] Here T(x) is the traffic density in bin x, wherein X is the
set of all bins in region 20, and p(S(A,x)) is the probability that
cell A serves mobile unit 23 in bin x. T(x) is a random variable,
which at this point is unknown, but is assumed to be non-negative.
An exemplary method for calculating p(S(A,x)) is described in the
above-mentioned provisional patent application. The sum of
p(S(Y,x)) over all cells Y in region 20 should be one (or zero in
uncovered bins).
[0039] In order to be able to estimate T(x), computer 37 also
receives one or more quality indicators collected from antennas 22
by MSC 36, at a quality measuring step 42. Preferably, the
following indicators are used:
[0040] Received power level statistics. For a given cell A, the
global received power level density is related to the local power
level density R(A,x) in each bin x by the expression: 2 p ( R ( R ,
X ) = b ) = 1 T ( A ) x X T ( x ) p ( S ( A , x ) ) p ( R ( A , x )
= b ) ( 2 )
[0041] R(A,x) is a random variable, preferably discrete-valued,
which represents the signal strength of cell A in bin x.
[0042] Handoff statistics 44. For a given cell A, the global
handoff density to any other cell, say cell B, is represented by H
(A.fwdarw.B,X), corresponding to the number of handoffs from cell A
to cell B per unit time over all of set X. Handoffs are coordinated
and monitored by MSC 36. The global handoff density is related to
the local handoff density H(A.fwdarw.B,x) in each bin x by the
expression: 3 p ( H ( A B , X ) ) = 1 T ( A ) x X T ( x ) p ( S ( A
, x ) ) p ( H ( A B , x ) ) ( 3 )
[0043] H(A.fwdarw.B,x) is a random variable, which depends on the
signal strengths of cells A and B in bin x and the criteria used in
the cellular network to decide when a handoff should take place.
Methods for calculating H are similarly described in the
above-mentioned provisional and regular patent applications.
[0044] Quality statistics 46. Each mobile unit 23 suffers from some
interference, resulting in a carrier/interference (C/I) value that
represents the strength of the carrier signal received by the
mobile unit from its serving cell, compared to the strength of the
interfering signals received from other cells in region 20 at the
same frequency. C/I, in other words, is a specific sort of
signal/noise ratio. The C/I ratio experienced by a mobile unit
determines the quality of its calls. The call quality is typically
measured in terms of quality parameters Q(A,x), such as BER (bit
error rate), FER (frame erasure rate) or RxQual (received level
quality), as mentioned above. The mobile units report their call
quality values to their serving cells. These values are aggregated
by MSC 36 to compute the global quality histogram for cell A,
Q(A,X), corresponding to the probability that a mobile unit served
by cell A anywhere in region 20 will measure some call quality b.
The global quality parameters are related to local quality variable
for each bin x, Q(A,x), by the expression: 4 p ( Q ( A , X ) = b )
= 1 T ( A ) x X T ( x ) p ( S ( A , x ) ) p ( Q ( A , x ) = b ) ( 4
)
[0045] Dropped calls. Each mobile unit served by the cellular
system and suffering from some interference may become subject to
the cellular system drop call mechanism. Cellular systems keep
record of drop rates of calls served by each cell. These dropped
call rates can thus be considered another form of quality
statistics. The global drop call parameters are related to a local
drop call variable D(A,x) for each bin x by the expression: 5 p ( D
( A , X ) ) = 1 T ( A ) x X T ( x ) p ( S ( A , x ) ) p ( D ( A , x
) ) ( 5 )
[0046] The calculations of drop probabilities take into account
channel allocation and technology-dependent mechanisms for dropping
calls.
[0047] Impact matrix 48. Each element of this matrix corresponds to
the interference probability between a pair of cells A and B,
assuming that both cells use the same frequency. In other words,
the matrix element IM(B.fwdarw.A,X) represents the percentage of
traffic served by cell A that would be damaged (typically by
reducing the C/I ratio to below some chosen threshold) due to
interference from cell B under such conditions. The impact matrix
elements for cell A can be determined by computer 37 based on
measurements made by mobile units in the area of cell A of the
relative signal strengths received from other cells. Such signal
strength data are commonly assembled by mobile units and reported
to MSC 36 for use in deciding when a given mobile unit should be
handed off to a new cell (mobile-assisted handoff). The impact
matrix elements may also be computed based on C/I statistics 46.
The global impact matrix elements are related to the local elements
IM(B.fwdarw.A,x) by the expression: 6 IM ( B A , x ) = 1 T ( A ) x
X T ( x ) p ( S ( A , x ) ) IM ( B A , x ) ( 6 )
[0048] In addition to one or more of these quality indicators,
computer 37 also receives timing advance statistics for each cell
in region 20, at a time measurement step 50. "Timing advance" is a
term used in GSM networks to refer to the delay t between the time
of transmission of a signal from antenna 22 and the time of its
reception by mobile unit 23 (or vice versa). Similar measurements
may be made in other types of mobile communication networks. The
time delay t is proportional to the distance d between the antenna
and the mobile unit. A terrain map is preferably used in
translating timing advance into distance from a site. Timing
advance measurements may thus be used to determine the distance
between the antenna 22 of the serving cell and the bin 38 in which
mobile 23 is located while served by the cell. We define the timing
advance variable TA(A,d) to be equal to the number of transmissions
received or transmitted in cell A during a given time period from
or to mobile units at distance d from the antenna. This variable is
related to the per-bin traffic density by the expression: 7 TA ( A
, d ) = ( x X | dist ( x , A ) = d ) T ( x ) p ( S ( A , x ) ) ( 7
)
[0049] In order to determine traffic density for different clutter
sub-types occurring in region 20, the region is divided up into
bins 38, at a binning step 52, as described above. The bins are
then grouped into different clutter sub-types, at a bin grouping
step 54. Various criteria may be used to define the sub-types
within a given clutter type, for example:
[0050] Region 20 may be divided by a grid, such as a
latitude/longitude or UTM grid. All bins 38 of a given clutter type
within the same square of the grid are defined as belonging to the
same sub-type.
[0051] Each bin 38 may be classified according to the best-serving
cell, i.e., the cell (or antenna 22) having the highest probability
of serving mobile units 23 in that bin (typically due to factors
such as antenna signal strengths and handoff parameters). All bins
of a given clutter type that belong to the same best-serving cell
are defined as belonging to the same sub-type.
[0052] Sets of mutually-adjacent bins 38 of the same clutter type
may be clustered together to define sub-types. Preferably, the size
of each set is limited by restricting the maximum distance between
any two bins in the set.
[0053] Alternatively, other criteria may be used to define
sub-areas and sub-types within region 20. The term "sub-type"
should therefore be understood to refer not only to areas having
different types of clutter characteristics, but more broadly to
encompass any classification of bins 38 that can be used to
differentiate areas and sub-areas by traffic density.
[0054] Computer 37 processes the global traffic statistics and
quality indicators for each cell in order to find the specific
traffic density for each clutter sub-type, at an analysis step 56.
The inputs to this calculation are the measured values of T(A),
along with one or more of H (A.fwdarw.B, X), R (A, X), Q(A, X), IM
(B.fwdarw.A, X) and TA (A, d), as measured for all cells A and B in
region 20. The measured values are inserted into equations (1)
through (5), as appropriate. The bin traffic density value variable
in each equation is replaced by the applicable sub-type traffic
density, i.e., T(x)=T(sub-type(x)). The set of equations thus
obtained is inverted to find T(sub-type(x)) for all the sub-types
chosen at step 54. The sub-type traffic densities are preferably
adjusted, if necessary, to maintain continuity of the local traffic
density among neighboring bins, since it is expected that the
traffic density will not change abruptly from one bin to the
next.
[0055] Once the traffic density for each clutter sub-type is known,
the density values can be mapped back to bins 38 according to their
respective sub-types. This mapping is typically used in optimizing
the operating configuration of antennas 22, at an optimization step
58. The frequencies allocated to the different cells in region 20
may be changed, based on the traffic density map, to give better
coverage in bins where there is dense traffic, while possibly
reducing wasted over-allocation in areas of sparse traffic. Other
factors, such as the height, transmission power, azimuth and tilt
of the antennas may also be adjusted, and extra antennas may be
added in problematic areas.
[0056] The computer program listing appendix to this application
contains a Microsoft Excel spreadsheet file, which illustrates
computer analysis of traffic statistics and quality indicators in
order to find specific sub-type clutter densities. The spreadsheet
file can be opened and operated using Excel version 2000 (Microsoft
Corporation, Redmond, Wash.), running on a personal computer with a
Pentium III processor and the Windows 2000 operating system. The
Excel "Solver" tool should be installed according to the
instructions provided with the spreadsheet software.
[0057] Upon opening the spreadsheet, the user will see a clutter
map at upper left, defining a number of different clutter types and
sub-types that are spread over a geographical area of interest. The
map is divided into a grid of 20.times.20 bins. The map layout can
be varied by changing the underlying numerical values. Below the
clutter map, at the left side of the spreadsheet, are power maps
showing power received from three different antennas, identified as
A, B and C, as a function of location. The received antenna powers
may similarly be modified by changing the underlying numerical
values. In actual operation, the values in the clutter and power
maps would typically be determined by values of these parameters
measured in the field or taken from existing maps and models.
[0058] At the right of the spreadsheet, a table of "switch-measured
values" contains values of traffic density (in Erlangs) served by
each of antennas A, B and C, as well as impact matrix and handoff
probability elements for each pair of the antennas. In actual
operation, these values would be derived from operational data
gathered by a cellular network switch serving the antennas. In the
spreadsheet, these values may be varied by the user. Further model
parameters to be input by the user are provided in the tables at
the upper right of the spreadsheet.
[0059] When the desired input values have been entered in the
tables, the user should select Tools >Solver in the Excel menu,
and should then click on the "Solve" button in the dialog box that
appears. The Excel Solver will compute the clutter density per
sub-type, and the computed values will appear in the clutter
density table at the upper right of the spreadsheet. The sub-type
clutter densities are calculated so as to minimize the differences
between the switch-measured values of the traffic density, impact
matrix and handoff probability (as input by the user) and the
corresponding values of these parameters that are derived from the
computational model. The model-derived parameters are calculated by
mapping the computed clutter densities back to the individual bins.
These calculations are performed iteratively until the Solver
reaches a solution within predetermined convergence limits. The
resulting traffic density, impact matrix elements and handoff
probabilities per bin are shown for each antenna in the maps in the
lower right-hand portion of the spreadsheet.
[0060] The techniques embodied in the attached spreadsheet may be
extended in a straightforward manner to larger and more complex
systems. Alternatively, other methods for solving sets of
constraints may be used in this context, as will be apparent to
those skilled in the art.
[0061] Although in the preferred embodiments described above,
certain particular quality statistics are used in building
estimates of traffic distribution, the principles of the present
invention are not limited to this set of statistical indicators.
Other quality measures that may be used in this context will be
apparent to those skilled in the art. In particular, while some of
the quality indicators measured at step 42 in the method of FIG. 2
are specifically characteristic of cellular voice communications,
the same method may easily be adapted for use in wireless packet
data networks. In such networks, switch statistics such as data
throughput and delay are routinely measured and can be used in
extracting traffic distribution information in a manner
substantially similar to that described above. Furthermore, whereas
certain of the quality statistics used in these preferred
embodiments are specific to narrowband cellular networks, the
principles of the present invention may also be applied to other
types of mobile communication networks, including broadband
cellular networks, such as CDMA-based systems.
[0062] It will thus be appreciated that the preferred embodiments
described above are cited by way of example, and that the present
invention is not limited to what has been particularly shown and
described hereinabove. Rather, the scope of the present invention
includes both combinations and subcombinations of the various
features described hereinabove, as well as variations and
modifications thereof which would occur to persons skilled in the
art upon reading the foregoing description and which are not
disclosed in the prior art.
* * * * *